Abstract
Although Data Envelopment Analysis (DEA) assumes that inputs and outputs take non-negative real values, in some real-world cases, data are integer-valued. In some situations, rounding a fractional value to the closest integer can lead to a misleading evaluation of efficiency and in some cases may lead to an infeasible projection point. To date, various radial and non-radial models have been presented. This paper proposes a slacks-based non-linear model that guarantees an integer-valued reference point for all integer targets. Also, the reference point of each target is feasible under the proposed model. The lack of a need to round answers to the closest whole value is an advantage of this method. In addition, the results of this model are compared with other models. An example is used to clarify the suggested method.
Original language | English |
---|---|
Pages (from-to) | 80-91 |
Number of pages | 12 |
Journal | Operations Research and Decisions |
Volume | 32 |
Issue number | 3 |
DOIs | |
Publication status | Published - 2022 |
Keywords
- data envelopment analysis
- integer-valued data
- non-radial model
- slack based measure
ASJC Scopus subject areas
- Statistics and Probability
- Modelling and Simulation
- Statistics, Probability and Uncertainty
- Management Science and Operations Research
- Management of Technology and Innovation